rates and radiocarbon content of summer ecosystem ... · the study period (2010, 2011, and 2012)...

Rates and radiocarbon content of summer ecosystemrespiration in response to long-term deeper snowin the High Arctic of NW GreenlandM. Lupascu1, J. M. Welker2, X. Xu1, and C. I. Czimczik1

1Department of Earth System Science, University of California, Irvine, California, USA, 2Department of Biological Sciences,University of Alaska Anchorage, Anchorage, Alaska, USA

Abstract The amount and timing of snow cover control the cycling of carbon (C), water, and energy in arcticecosystems. The implications of changing snow cover for regional C budgets, biogeochemistry, hydrology, andalbedo due to climate change are rudimentary, especially for the High Arctic. In a polar semidesert of NWGreenland, we used a ~10 year old snow manipulation experiment to quantify how deeper snow affectsmagnitude, seasonality, and 14C content of summer C emissions. We monitored ecosystem respiration (Reco),soil CO2, and their

14C contents over three summers in vegetated and bare areas. Additional snowpack, elevatedsoil water content (SWC), and temperature throughout the growing season in vegetated, but not in bare, areas.Daily Reco was positively correlated to temperature, but negatively correlated to SWC; consequently, we foundno effect of increased snow on daily flux. Cumulative summertime Reco was not related to annual snowfall, butto water year precipitation (winter snow plus summer rain). Experimentally increased snowpack shortened thegrowing season length and reduced summertime Reco up to 40%. Soil CO2 was older under increased snow.However, we found no effect of snow depth on the Reco age because older C emissions were masked byyounger CO2 produced from the litter layer or plant respiration. In the High Arctic, anticipated changes inprecipitation regime associated with warming are a key uncertainty for understanding future C cycling. In polarsemideserts, water year precipitation is an important driver of summertime Reco. Permafrost C is vulnerable tochanges in snowpack, with a deeper snowpack-promoting decomposition of older soil C.

1. Introduction

The High Arctic (>70°N) is undergoing rapid warming [Overpeck et al., 1997] and “greening” [Bhatt et al., 2010;Epstein et al., 2013]. Here polar semideserts cover approximately 1 × 106 km2 and, together with deserts andmires, hold an estimated 12 pg of organic carbon (C) in soils affected by near-surface permafrost [Burnhamand Sletten, 2010]. These soil C pools are exceedingly heterogeneous in their vertical and surface traits[Horwath et al., 2008], as differential frost heave and patterned ground result in extensive vertical mixing anda surface that is comprised of vegetated and bare areas [Kessler and Werner, 2003; Walker et al., 2004]. Field-based studies investigating what controls C losses and sequestration in the complex soils of the High Arcticare important to fully understanding C dynamics in permafrost soils globally [Welker et al., 2004].

A key driver of C cycling in arctic ecosystems is the onset, duration, and amount of snow cover. Arctic tundrais covered in snow for 8–10 months of the year, although spatial and interannual variability of snow cover ishigh [Fahnestock et al., 1999; Jones et al., 1999; Callaghan et al., 2011]. Anticipated changes in precipitation,including the amount and timing of snow cover as a consequence of climate warming, and their implicationsfor regional biogeochemistry (i.e., C cycling), albedo, and hydrology are highly uncertain [Hinzman et al., 2005;Steffen, 2006]. Global climate models (GCM) projection in snow cover for 2050 indicates increases inmaximum soil water equivalent (SWE) by up to 15% for most of the panarctic, with the greatest increase(15–30%) over Siberia. On the other hand, GCM project a decrease in snow cover duration (SCD) by ~10–20%for most of the panarctic by 2050, with the smallest decreases over Siberia (

Due to its high albedo and low thermal conductivity, the depth and timing of snow cover control tundrawater and energy budgets [Pomeroy et al., 2006; Adam et al., 2009], active layer thermal regime [Lafrenièreet al., 2013], and permafrost stability [Lawrence et al., 2008]. Snow cover also influences ecosystembiogeochemistry, such as the cycling of C and nutrients. These effects are mediated via changes in (1)growing season length that in turn affects plant productivity, phenology, reproductive success, and tissuecomposition [Galen and Stanton, 1995; Rieley et al., 1995; Molau and Edlund, 1996; Welker et al., 2000, 2005;Groendahl et al., 2007; Borner et al., 2008; Bhatt et al., 2010; Cooper et al., 2011; Olofsson et al., 2011; Leffler andWelker, 2013; Epstein et al., 2013] and (2) soil microbial activity [Schimel et al., 2004; Sturm et al., 2005; Borneret al., 2008]. Associated changes in winter nutrient cycling can have subsequent effects on annualC budgets [Welker et al., 2000], culminating in shifts in tundra vegetation composition and canopy structure[Sturm et al., 2005; Wahren et al., 2005; Tape et al., 2006; Forbes et al., 2010], which, in turn, feedback to howsnow accumulation is distributed across the landscape [Sturm et al., 2005].

Several studies have explored the effects of increased winter snow depth on soil and ecosystem C cycling.Deeper snowpack typically results in greater wintertime soil temperatures [Jones et al., 1998;Walker et al., 1999;Schimel et al., 2004; Welker et al., 2004; Hinkel and Hurd, 2006] and ecosystem respiration (Reco) [Brooks et al.,1997, 1998;Walker et al., 1999;Welker et al., 2000; Schimel et al., 2004;Morgner et al., 2010; Nobrega and Grogan,2013]. However, in both the High and Low Arctic effects on summertime C cycling are unclear, with somestudies exhibiting an increase in Reco under experimentally increased snowpack [Natali et al., 2011; Rogers et al.,2011], a decrease [Jones et al., 1998;Welker et al., 2000; Björkman et al., 2010] or no effect [Björkman et al., 2010;Morgner et al., 2010]. Furthermore, only two studies, both carried out in Alaska [Nowinski et al., 2010; Natali et al.,2011], have investigated the effects of winter snow manipulation on soil C storage and turnover.

Here we investigated the effects of long-term (10 years) increases in winter snowpack on the magnitude,seasonality, and sources of summertime Reco. We monitored the rate of Reco and the concentration of CO2 inthe soil pore space along with their radiocarbon (14C) content; nondestructive techniques used to inferchanges in ecosystem and soil C sequestration or loss [Trumbore, 2006, 2009]. All living organisms arelabeled with 14C, which is produced in the atmosphere, oxidized to CO2, and enters the food chain viaphotosynthesis. Testing of thermonuclear bombs aboveground in the midtwentieth century increased theamount of atmospheric 14CO2 above natural production levels. The amount of this bomb

14C in theatmosphere is declining due to mixing with terrestrial and ocean C pools and dilution by fossil (14C-free)CO2 [Levin et al., 1980]. Consequentially, the

14C content of recent photosynthetic products as well as that ofplant and rhizosphere respiration is similar to the 14C content of current atmospheric CO2. The respiration ofsoil microbes decomposing organic matter made from photosynthetic products years to decades ago isenriched in 14C. Furthermore, microbial respiration decomposing hundreds to thousands or years old organicmatter is depleted in 14C due to radioactive decay. Any increases in decomposition of old permafrost C wouldhave important consequences for atmospheric CO2. Decomposition of old C that was not part of the active Ccycle for millennia results in an increase flux of C to the atmosphere while rapid cycling of young C betweenplants and microbes has a near-zero effect on the atmospheric C pool [Trumbore, 2009]. Specifically, with thisstudy we asked what are the effects of increasing winter snowpack on (1) the concentration and 14C content ofCO2 within the active layer and (2) on the magnitude, seasonality, and

14C content of Reco during the summer.

2. Material and Methods2.1. Site Description

The study was carried out in prostrate dwarf-shrub tundra in the High Arctic of northwest Greenland nearThule Air Base (76°32′N, 68°50′W; 200–350mabove sea level) [Sullivan et al., 2008; Sharp et al., 2013]. Meanannual air temperature is�11.3± 1.3°C, with 122.6±45.4mmofmean annual precipitation (1952–2012). Duringthe study period (2010, 2011, and 2012) mean summer temperature was 5.6±2.0°C, precipitation was24.0± 18mm, andmean winter snowfall 1306±200mm (data from Thule (THU) airport). During the last climatenormal period (1983–2012), air temperature shows an increasing trend of 1.03°C/decade with the strongestwarming during the winter months (December, January, and February) of 1.84°C/decade [Lupascu et al., 2014b].

The vascular plant community at our site is dominated by the deciduous dwarf-shrub Salix arctica PALL., thegraminoid Carex rupestris ALL., and the wintergreen dwarf-shrub Dryas integrifolia VAHL. The live biomass and

Journal of Geophysical Research: Biogeosciences 10.1002/2013JG002494

LUPASCU ET AL. ©2014. American Geophysical Union. All Rights Reserved. 2

litter of these three species account for approximately 70% of vascular plant cover. The soil is a TypicHaploturbel [U.S. Department of Agriculture, 1999] with a thin, patchy litter layer, high-gravel content, and,based on soil pits near the experimental area, a maximum thaw depth of about 1.0 ± 0.2m [Sullivan et al.,2008, Horwath, 2007]. Soils are well drained and subject to cryoturbation. In vegetated areas, the topsoil(0–12 cm) texture is 67–74% sand, 20–34% silt, and 5–8% clay, while in bare areas the texture is 54–64% sand,33–38% silt, and 3–7% clay [Sullivan et al., 2008]. For the top 40 cm, soil bulk density is approximately1.10 g cm�3 and the organic soil C content is 0.2–1.6% for vegetated areas and 0.1–0.2% for bare areas.

2.2. Experimental Set Up

Measurements were conducted at a long-term snowpack manipulation experiment consisting of two 1.2 talland 6.0m long snow fences erected in the summer of 2003 [Rogers et al., 2011]. Fences were alignedperpendicular to the dominant winter wind direction, and snow naturally accumulates on the leeward side in acontinuously tapering drift. At each fence, 2.0 × 0.8m2 plots were delimited within three larger, preestablishedmain plots at each of three levels of winter snowpack: ambient (control, ~0.25m), intermediate ((Snowfencemedium) SFM, ~0.55m), and deep ((Snowfence high) SFH, ~1.1m) [Rogers et al., 2011]. Ambient plots used inthis study are the same control plots used in the factorial warming× irrigation experiment described by Lupascuet al. [2014a, 2014b] and Sullivan et al. [2008]. Plots were oriented to span the transition between vascular plantsand bare soil/cryptogamic crust, such that each comprised approximately 50%, to facilitate scaling from the plotto ecosystem level. Thus, each plot had two subplots representing the vegetated versus bare areas (Figure S1 inthe supporting information). Direct measurements of snow depth were not made as part of this study, as thesites could not be accessed during winter. Measurements were taken during the snow-free period, fromwhen itwas possible to insert chamber bases and conduct Reco measurements without compacting snow (from May/early June under ambient conditions to 20 August in 2010, 2011, and 2012).

2.3. Weather Data

Weather data were calculated based on temperature, precipitation, and snowfall data from the Thule airport(THU) weather station for the period 2009–2012. Daily mean temperatures are calculated as the mean of thedaily minimum and maximum temperatures. Mean summer temperatures represent the average of the dailymean temperature for the June–August period. Cumulative precipitation represents the sum of the dailyprecipitation (rain, mm) for the summer (June, July, and August). Cumulative snowfall is the sum of the dailysnowfall from September to May.

Water year precipitation, a concept commonly used in hydrology to describe bioavailable precipitation in(semi-) arid regions, is the sum of precipitation and snowfall (expressed as SWE). We calculated SWE as theaverage of the cumulative snowfall times its minimum (252 kgm�3) or maximum (354 kgm�3) density, usingdensity values reported for high flat areas in the High Arctic [Woo et al., 1983].

2.4. Ecosystem Respiration and Soil Pore Space CO2 Concentrations

Ecosystem respiration and pore space CO2 concentrations were measured once per day during 9 A.M. to 12noon, 2 to 3 times per week. Ecosystem respiration was determined through the use of opaque chambers(30 cm internal diameter; 8 L volume). In June, collars were inserted to ~ 2 cm depth, sealed with soil materialon the outside, and left in place for the sampling season. To calculate Reco, air was circulated between thechamber’s headspace connected to an infrared gas analyzer and a data logger (LI-840, LI-1400, and LI-CORBiosciences, Lincoln, NE, USA) at a rate of 0.5 Lmin�1. Flux rates were estimated from the slope of time versusCO2 concentration curves using linear regression. In parallel, we measured soil temperature (15–077, FisherScientific, resolution ±0.1°C) at 5 and 10 cm depth below the soil surface and soil water content (SWC,Hydrosense, Campbell Scientific, Logan, UT, USA, resolution ±0.1%) at 5 cm depth. Direct measurements ofactive layer depth could not be made due to the small plot size and high-gravel content. However, weestimated thaw progression using our gas wells (see below), as soil gas cannot be retrieved from frozen soils.Rarely, soil gas could not be retrieved from wells, because soils were temporarily water logged duringsnowmelt or following rain storms.

Carbon dioxide concentrations in the soil pore space were analyzed via stainless steel gas wells (0.35 cm ID,0.6 cm OD) from 20, 30, 60, or 90 cm soil depth. Wells were inserted in 2010 and 2012, capped with rubbersepta (Blue septa, Grace, Deerfield, IL, USA), and left in the ground during the entire study period including



the winter (n= 1–2). Gas samples were taken with a 60mL syringe (BD, Franklin Lakes, NJ, USA) and injected(first sample extraction was always discarded) into an infrared gas analyzer connected to a data logger(LI-800, LI-1400, and LI-COR). Concentrations were recorded manually.

2.5. Estimating Summer Ecosystem Respiration Fluxes

To estimate summer Reco, we used the relationship between respiration and temperature [Lloyd and Taylor,1994], represented by an Arrhenius type equation (1), where the effective activation energy for respirationvaries inversely with temperature.

Reco ¼ Ae�EoT�Toð Þ (1)

with (Eo) = 308.56 K and (To) = 227.13 K and where (A) is a data set-dependent variable. (A) was first obtainedusing the data collected in situ, and then daily Reco for the missing days was calculated using the averagedaily air temperature (T ). Daily Reco for the growing season (snow-free period) was then summed up to yieldcumulative summertime Reco.

2.6. Gas Sampling for Isotope Analyses

To sample Reco for14C analysis, chambers were left closed until the CO2 concentration inside the chamber

was about twice that in ambient air (up to 48 h). After measuring the CO2 concentration inside the chamberheadspace, CO2 was collected by circulating the air inside the chamber through drierite (W. A. HammondDrierite Co. Ltd., Xenia, OH, USA) followed by a preconditioned, activated molecular sieve (powder-free133 8/12 beads, Grace) trap at a rate of 0.5 Lmin�1 for 15min [Gaudinski et al., 2000]. During eachsampling event, we also collected CO2 in ambient air on molecular sieve traps.

Pore space CO2 was collected in evacuated stainless steel canisters via flow-restricting stainless steelcapillaries (0.010 × 0.063 × 30 cm, Fisher Scientific, Pittsburgh, PA, USA) to minimize disturbing the soil CO2concentration gradient and sampling air from other than the sampling depth [Gaudinski et al., 2000]. Duringeach sampling event, we also collected CO2 in ambient air on molecular sieve traps.

Carbon dioxide respired from roots in the control plot were sampled by manually extracting all roots from acore of 30 × 5 cm the same day the soil was sampled. Roots were rinsed with water and placed into a 1 Lmason jar, which was flushed with CO2-free air. After 24 h, the CO2 produced was collected on a molecularsieve trap. To conserve the integrity of this long-term experiment, roots could only be retrieved from controlareas outside the experimental plots.

2.7. Isotope Analyses of CO2

Carbon dioxide was released from molecular sieve traps by baking at 650°C for 45min or extracted fromcanisters using a vacuum line, purified cryogenically, and reduced to graphite via Zn reduction [Xu et al.,2007]. A split of the CO2 was analyzed for its δ

13C ratio (GasBench II, DeltaPlus, Thermo Scientific). The 14Ccontent of the graphite was measured with accelerator mass spectrometry (NEC 0.5MV 1.5SDH-2 acceleratormass spectrometry (AMS)) at the Keck Carbon Cycle (KCC) AMS laboratory of the University of California (UC)Irvine [Southon and Santos, 2007]. Data are reported relative to National Institute of Standards andTechnology OX-I (SRM 4990a) and OX-II (SRM 4990c) [Stuiver and Polach, 1977]. Themeasurement uncertaintyfor Δ14C was < 2 per mil.

The 14C content of Reco was corrected for the amount of CO2 from ambient air present in each chamber’sheadspace:

Δ14Ccor: ¼Δ14Cobs: � f air � Δ14Cair

� �

1 � f airð Þ (2)

with (Δ14Ccor.) being the actual14C content of Reco, (Δ

14Cobs.) the measured14C content of a given sample

and fair the fraction of CO2 derived from ambient air, calculated from the CO2 concentrations inside thechamber immediately before trapping and in ambient air.

2.8. Statistical Analyses

Control plots had a sampling size of n=2, while we increased the sampling size of treatment plots from onein 2010 to two in 2011 to four in 2012. Our sample for vegetated areas consisted of 662 observations for Reco,



with the control plot as experimental unit. We conducted a two-way analysis of variance (ANOVA) to examinethe fixed effects of measurement year (2010, 2011, and 2012), treatment type (control, SFM, and SFH) and theinteraction of the two on Reco. The same analysis was applied for SWC (427 observations), soil temperature(393 observations). Compared to Reco fluxes, in the

14C analyses, the extra variable “plot type” (vegetatedversus bare areas) was taken into account and 96 observations were used for the ANOVA.

3. Results3.1. Weather Conditions

The three sampling years exhibited contrasting weather. Average summer air temperature was highest in2011 (~6.1 ± 3.1°C), with the maximum difference observed in July (+1–2°C) (Figure 1a). Summertimetemperature was similar in 2010 and 2012 (~5.3 ± 3.0°C; Figure 1a). In contrast, cumulative summerprecipitation was highest in 2012 (161.3mm) compared to 2010 (84.8mm) or 2011 (36.7mm) (Figure 1f). Wealso observed a striking difference in the timing of summer precipitation. In 2010 and 2011 (Figures 1b and1c), more than half of summer rain fell during the month of July (61–73%) while June was very dry (1.1–6.5%).In 2012 (Figure 1d), precipitation was more evenly distributed, with 31, 46, and 22% of the total summerprecipitation falling in June, July, and August, respectively. Cumulative snowfall was highest in winter2010/2011, with +36% more snow compared to 2009/2010 and +15% compared to 2011/2012 (Figure 1e). Ifwe consider the water year precipitation, 2012 was still the wettest year with 557 ± 67mm, followed by 2011(492 ± 77mm) and 2010 (410 ± 56mm).

3.2. Soil Water Content and Temperature

Surface SWC in vegetated areas (Figure 2, middle) was strongly influenced by interannual difference in snowcover, summer precipitation patterns, and experimental treatment. The water year precipitation affected SWCthroughout the growing season in all years with SWC being greater in 2012 than in 2011 or 2010. Soil watercontent increased during spring due to snowmelt and again in July and August due to rainfall. We observedthe highest average summer SWC in 2012 (31.8 ± 3.7 vol %), compared to 2010 and 2011 (15.1 ± 4.4 and17.2 ± 4.4 vol %, respectively) (year difference, F=36.88, p< 0.001). Higher experimental snowpack levelincreased average summer SWC by +4.1 ± 0.5 (SFM) and +5.2 ± 1.9 vol % (SFH) relative to the ambient snowconditions but did not alter the seasonal pattern (F=9.90, p< 0.002). Soil water content in the bare areas didnot show any statistical difference between deep and ambient snow areas (data not shown).

Surface soil temperature (Figure 2, top) reflected the seasonal change in air temperature, with maximumvalues in July, as previously observed [Czimczik and Welker, 2010]. In vegetated areas, experimentally increasedsnowpack affected soil temperatures during all summers. Compared to the control (mean 8.6 ± 0.3°C),experimental plots (SFM and SFH) were +0.9 ± 0.2°C warmer (F= 7.055, p< 0.024). No statistical difference

Figure 1. (a) Summer daily mean air temperature and (b–d) daily precipitation (rain) patterns for Thule, Greenland, (Thuleairport, THU) for 2010–2012. (e) Cumulative snowfall for the preceding winter (2010 =winter 2009/2010) and (f) cumulativesummer precipitation for 2010–2012.



between the two snowpack treatments was found. On the other hand, a distinct difference between the years(F=6.112, p< 0.002) was observed, with soil temperatures being the warmest in 2011 (mean 10.0± 0.3 °C),the year with the most snow.

3.3. Growing Season Length

The length of the growing season (i.e., snow-free period) varied between years due to the different snowpacklevel, with more snowfall resulting in a shorter growing season. The growing season was longest in 2010with 81 days in the control compared to 71 and 72 days in 2011 and 2012, respectively. In the treatments,growing season length was shorter than in the control, starting about 6 days later in SFM (medium snowpackplots) and about 11 days later in SHM (high snowpack plots) (Table 1).

3.4. Ecosystem Respiration

Summer ecosystem respiration was noticeably influenced by the seasonal variability in precipitation andtemperature, experimental snowpack level, and water year precipitation. On a daily scale, all plots andtreatments showed a weak negative correlation between daily Reco and SWC (R

2 = 0.22 ± 0.19), but a robustpositive correlation with soil temperature (R2 = 0.56 ± 0.19). Furthermore, Reco diminished considerablyduring cold spells, which typically concurred with rainfall events.

We observed two different seasonal patterns of Reco fluxes. In 2011 and 2012, Reco fluxes in thevegetated areas showed a sharp peak during snowmelt, and then followed the seasonal changes in airtemperature, with a maximum in July (Figures 2 (bottom) and 1a). Midsummer fluxes were of similarmagnitude in 2011 and 2012 (control, mean ± standard deviation: 2.36 ± 0.18 μmol Cm�2 s�1). In 2010,we did not observe a snowmelt peak as our measurements commenced a couple of weeks aftersnowmelt (DOY 151), and Reco fluxes could be divided into two distinct periods. Ecosystem respirationwas low and stable (control: 0.28 ± 0.11 μmol Cm�2 s�1) from early June (DOY 152) to mid-July (DOY195), and greater from mid-July to the end of August (control: 0.60 ± 0.29 μmol Cm�2 s�1; Figure 1b). Inbare areas, Reco fluxes were lower than in vegetated areas (

In all treatments, cumulative summertime Reco showed a positive trend with total water year precipitation,but not with winter snowfall. Summertime Reco was lowest in 2010 (Figure 3)—the year with the lowest wateryear precipitation. Additional water year precipitation in 2011 and 2012 stimulated summer Reco (Figure 3;year difference in control, F= 74.17, and p< 0.001). However, this increase in summer Reco was muted in thesnowpack manipulations.

In the snowpack manipulations, daily Reco fluxes in vegetated areas were similar to those in the control(Figure 2, bottom). However, due to a shorter snow-free period, cumulative summertime Reco fluxes in theSFM and SFH treatments were reduced by �29.8 ± 6.1 and �22.6 ± 27.2%, respectively, compared to thecontrol (F=16.42, p< 0.001; Figure 3). In bare areas, however, daily Reco fluxes in the SFM and SFH treatmentswere greater or similar to (Table 1) to the control Reco in 2010 and 2011, but lower in 2012. We found nostatistical differences in the cumulative summertime Reco fluxes between the two snowpack treatments.

3.5. Soil Pore Space CO2

Pore space CO2 concentrations reflect both production at the measured depth and diffusion within the soil.Concentrations were strongly affected by water year precipitation, seasonal changes in active layer depth, and

Table 1. Ecosystem Respiration (Reco) Fluxes From Vegetated and Bare Areas and From the Landscape (Vegetated+Bare)During Three Sampling Years (2010–2012)

Daily Flux (ave. ± SE, n=1–4) Monthly Flux

Vegetated Bare Landscape Vegetated Bare Landscape

Months (g Cm�2 d�1) Daysa (g Cm�2 d�1)

June 2010Control 0.24 ± 0.00 0.04 ± 0.03 0.28± 0.00 30 7.3 ± 0.1 1.1 ± 0.7 8.4 ± 0.7SFM 0.23 ± na 0.05 ± na 0.29 ±na 22 5.3 ± na 1.4 ± na 6.7 ± naSFH 0.12 ± na 0.07 ± na 0.27 ±na 17 2.1 ± na 1.2 ± na 3.3 ± na

July 2010Control 0.53 ± 0.01 0.07 ± 0.05 0.60± 0.00 31 16.5 ± 0.2 2.0 ± 1.4 18.5 ± 1.4SFM 0.40 ± na 0.10 ± na 0.50 ±na 31 12.5 ± na 3.0 ± na 15.5 ± naSFH 0.36 ± na 0.11 ± na 0.48 ±na 31 11.3 ± na 3.6 ± na 14.9 ± na

August 2010Control 0.61 ± 0.02 0.09 ± 0.06 0.69± 0.00 20 12.1 ± 0.5 1.8 ± 1.2 13.9 ± 1.3SFM 0.47 ± na 0.12 ± na 0.59 ±na 20 9.3 ± na 2.5 ± na 11.8 ± naSFH 0.42 ± na 0.13 ± na 0.54 ±na 20 8.3 ± na 2.5 ± na 10.8 ± na

June 2011Control 0.86 ± 0.12 0.12 ± 0.07 0.98± 0.14 20 17.2 ± 2.3 2.4 ± 1.4 19.6 ± 2.7SFM 0.61 ± 0.07 0.14 ± 0.01 0.76± 0.07 14 8.6 ± 1.0 2.0 ± 0.1 10.6 ± 1.0SFH 1.14 ± 0.03 0.15 ± 0.01 1.07± 0.04 9 10.2 ± 0.5 1.4 ± 0.2 11.6 ± 0.5

July 2011Control 1.43 ± 0.09 0.22 ± 0.13 1.65± 0.16 31 44.4 ± 2.7 6.9 ± 4.1 51.3 ± 4.9SFM 1.11 ± 0.17 0.26 ± 0.00 1.36± 0.17 31 34.4 ± 5.3 7.9 ± 0.1 42.3 ± 5.3SFH 1.71 ± 0.04 0.22 ± 0.02 1.93± 0.04 31 53.0 ± 1.2 6.8 ± 0.5 59.8 ± 1.7

August 2011Control 1.42 ± 0.11 0.22 ± 0.13 1.64± 0.17 20 28.4 ± 2.2 4.4 ± 2.5 32.8 ± 3.3SFM 0.85 ± 0.10 0.20 ± 0.01 1.04± 0.07 20 16.9 ± 2.1 3.9 ± 0.2 20.8 ± 2.1SFH 1.29 ± 0.01 0.16 ± 0.01 1.46± 0.01 20 25.9 ± 0.2 3.3 ± 0.1 29.2 ± 0.2

June 2012Control 1.43 ± na 0.33 ± na 1.76 ±na 21 30.1 ± na 7.0 ± na 37.1 ± naSFM 0.97 ± 0.06 0.23 ± 0.11 1.20± 0.06 16 15.5 ± 0.9 3.7 ± 1.8 19.2 ± 2.0SFH 1.08 ± 0.05 0.18 ± 0.05 1.26± 0.04 11 11.8 ± 0.8 2.0 ± 0.7 13.8 ± 1.1

July 2012Control 1.99 ± na 0.48 ± na 2.46 ±na 31 61.6 ± na 14.8 ± na 76.4 ± naSFM 1.36 ± 0.08 0.34 ± 0.17 1.70± 0.18 31 42.2 ± 2.4 10.4 ± 5.2 52.6 ± 5.7SFH 1.41 ± 0.09 0.20 ± 0.07 1.61± 0.12 31 43.6 ± 2.9 6.3 ± 2.1 49.9 ± 3.6

August 2012Control 1.74 ± na 0.41 ± na 2.15 ±na 20 34.7 ± na 8.2 ± na 42.9 ± naSFM 1.08 ± 0.04 0.26 ± 0.12 1.34± 0.08 20 21.6 ± 0.8 5.2 ± 2.4 26.8 ± 2.5SFH 1.20 ± 0.09 0.16 ± 0.04 1.37± 0.07 20 24.1 ± 1.9 3.3 ± 0.9 27.4 ± 2.1

aActual days of study period representing the growing season length; na = data not available (n=1).



snowpack manipulation (Figure 4). Vegetated areasshowed overall greater CO2 concentrations comparedto the bare areas (unpaired t test, p< 0.005), especiallyin the surface soil. Typically, CO2 concentrationspeaked between 30 and 60 cm, followed by adecrease toward the permafrost table (Figure S3).

During the summer, CO2 concentrations in thesurface soils exhibited a bimodal trend: A first peakafter snowmelt (up to 18000 ppm in vegetatedareas), and a second peak usually coupled withmaximum air temperatures and active layer depth(Figure 4). The magnitude of the snowmelt peakbetween years was positively related withsnowpack, with 2010 being the lowest (Figure 1e).However, interannual variability in theconcentration of pore space CO2 was smaller thanthat of Reco. The timing of the snowmelt and theassociated CO2 peak were delayed in years withgreater snowfall and in the snowpacktreatments (Figure 4).

Snowpack manipulation significantly increased CO2 concentrations at all depths compared to the control(Figure 4). For example, in 2012 seasonally averaged CO2 concentrations in the SFM treatment were4037± 903, 4660 ± 628, and 4418 ± 732 ppm at 20, 60, and 90 cm versus 2150 ± 187, 3500 ± 299, and3481± 432 ppm in the control (seasonal average ± SE; p< 0.002). In 2010 and 2011, we found no difference in

Figure 3. Cumulative summertime ecosystem respiration(Reco) from High Arctic tundra during 2010–2012 underambient (control), medium (SFM), and high (SFH) snowpackconditions (average± SE, n=1–4) (The number on the top ofeach histogram represents the growing season length cal-culated from the span of the actual measurement period.).

Figure 4. Pore space CO2 concentration at 20 and 60 cm depth in vegetated areas under ambient (control), medium (SFM), and high (SFH) snowpack conditionsduring the summer of 2010, 2011, and 2012 (n=1–2 per treatment). Brick- and zigzag-patterned histograms represent the presence of snow at the time of mea-surements in ambient and snow fence plots, respectively. (Note difference in scale between concentrations at different depths.)



the CO2 concentration between the two snow depth manipulations. In 2012, CO2 concentrations werehighest in the SFH treatment.

3.6. Radiocarbon Content of Ecosystem Respiration

The 14C content of Reco was significantly different between vegetated and bare areas and interannually butshowed only negligible differences seasonally or between control and experimental snow depthmanipulations. In vegetated areas, Reco was significantly enriched (younger) in

14C compared to Reco in bareareas in both the control and snowpack manipulations (F=1.16, p< 0.007; Figure 5). For example, theaverage summertime 14C values of Reco in the SFH treatment in 2010–2012 were 43.8 ± 2.0, 30.7 ± 1.7, and54.2 ± 2.2 per mil, respectively, in vegetated areas, and 23.2 ± 0.7, �18.8 ± 12.8, and 18.8.6 ± 5.4 per mil,respectively, in bare areas.

Variability in weather conditions between the three sampling years significantly affected the mean age ofReco (F=16.63, p< 0.001), with the most

14C depleted (oldest) Reco observed in the warmest year 2011 forboth vegetated and bare areas (Figure 5). For example, the average summertime 14C value of Reco invegetated areas of the SFM treatment was 19.9 ± 9.2 per mil in 2011, compared to 7.6 ± 2.0 per mil in 2010and 55.9 ± 2.2 per mil in 2012.

There was no seasonal trend in the 14C content of Reco in any year (Figure 5). Sporadically, however, wedetected very depleted 14C values (up to �190.8 ± 7.1 per mil) indicative of old Reco, which appeared to becoupled to rain pulses in July and August [Lupascu et al., 2014b]. We found no statistical difference betweenthe 14C values of Reco across the three snow treatment conditions (F=2.03, p< 0.158).

3.7. Radiocarbon Content of Soil Pore Space CO2

The 14C content of CO2 in soil gas, sampled below the rooting zone, provides insight to changes in microbialactivity and soil physical properties [Phillips et al., 2013; Torn et al., 2013]. The 14C values of pore space CO2exhibited two different depth trends (Figure 6): (1) An increase in age with depth or (2) an increase toabout 60 cm followed by a decrease toward the permafrost table. Overall, the 14C values of pore space CO2were older in bare than in vegetated areas (unpaired t test, p< 0.007).

Experimental increases in snow affected the 14C content of pore space CO2 throughout the soil profile, withdeeper snow resulting in older C at all depths. For example, the summertime average 14C content of soil CO2at 60 cm in bare areas was �14.0±11.4 and �67.7±11.5 per mil for SFM and SFH, respectively, compared to

Figure 5. Radiocarbon content of ecosystem respiration (Reco ± analytical error, n=1) from (a) bare and (b) vegetated areasunder ambient (control), medium (SFM), and high (SFH) snowpack conditions during the summers of 2010–2012 (Coarsearea indicate the range of 14CO2 in the atmosphere for 2010–2012; dotted line represent the litter layer; short-dash line,dash-dot-dot line, and long-dash line represent root incubation from the control plot in 2010, 2011, and 2012, respectively;circled symbols represent old C pulses concurring with small rain event (

27.9 per mil in the control. Two exceptions to this pattern were observed: pore space CO2 in the control wasolder than in the treatments in bare areas in 2011 (Figure 6a), and in vegetated areas at 90 cm in 2012(Figures 6d and 6e). For the latter event, sampling of CO2 from the snowpack manipulations, but not from thecontrol, coincided with intense rain events (>8mm). Within a given snow depth treatment, we found nosignificant seasonal or interannual differences in the 14C values (ages) of pore space CO2.

4. Discussion4.1. Effects of Snowpack on SWC and Soil Temperature

In vegetated areas we found that greater winter snowfall and experimentally increased snow depth resultedin greater summer SWC and soil temperatures throughout the growing season. While some studies havereported similar results to ours [Chimner and Welker, 2005;Morgner et al., 2010; Rogers et al., 2011], others didnot find any increase in SWC over the summer [Buckeridge and Grogan, 2010]. This is not surprising as soil(porosity and water-holding capacity) and vegetation type can play a crucial role in affecting SWC [De Micheleet al., 2008]. Furthermore, depending on the photosynthesis rates, plants can use more or less water [Sullivanand Welker, 2007] thus affecting SWC. In our study area, the prolonged higher SWC is likely due to greaterwater retention in the litter layer and rhizosphere as bare areas drained much faster.

It was previously shown elsewhere [Welker et al., 2000; Schimel et al., 2004; Morgner et al., 2010] that deepersnowpack increases soil temperature during the winter, but summertime effects are more complex. Whilemany studies showed lower summertime temperatures in areas with higher snowpack [Walker et al., 1999;Schimel et al., 2004; Hinkel and Hurd, 2006], others found no difference or warmer soil temperatures [Joneset al., 1998; Rogers et al., 2011]. These contrasting observations can be reconciled by differences in watercontent, vegetation cover, and increased litter fall [Fahnestock et al., 2000]. As water has a higher heat capacitythan air, soil with a greater SWC gain and loose heat more slowly than drier soils [Al-Kayssi et al., 1990].

Figure 6. Radiocarbon content of pore space CO2 along the soil profile under ambient (control), medium (SFM), and high (SFH) snowpack conditions in (a, b) bareand (c, d) vegetated areas during 2011 and 2012 (average± SE, n=1–3) (Dotted lines indicate the range of 14CO2 in ambient air.). (e) Concentration of pore space CO2at 90 cm and precipitation during the summer of 2012. (Circle symbols with dotted lines represent the timing of 14C samples for each treatment.)



In addition, soils and vegetation types with a better water retention capacity also minimize energydissipation by latent heat flux, further increasing soil warming. Greater leaf litter deposition associatedwith deeper snow may further insulate the soil surface from the atmosphere [Fahnestock et al., 2000].

4.2. Seasonality and Magnitude of Ecosystem Respiration

Interactions of air (and tightly coupled soil) temperature and precipitation (and hence SWC) are key drivers ofsummertime Reco fluxes in polar semideserts, as recently demonstrated in a summertime precipitation andtemperature manipulation experiment adjacent to the study reported here [Lupascu et al., 2014a]. On a dailytimescale, Reco fluxes were driven by temperature and showed a peak inmid-July that coincided withmaximumair temperatures and active layer depth. Mean daily summertime Reco fluxes observed in this study underambient snow are similar to those seen in other studies of High Arctic prostrate dwarf-shrub and herb tundraecosystems [Jones et al., 2000; Lloyd, 2001;Welker et al., 2004; Björkman et al., 2010; Morgner et al., 2010; Rogerset al., 2011]. On a seasonal timescale, the magnitude of total summertime Reco was controlled by the amount ofavailablewater year precipitation. Thus, changes in the amount of both summertime rainfall and snow cover arecritical uncertainties in understanding future C cycling in rapidly warming polar semideserts.

We found no effect of experimental increases in snow depth on daily Reco fluxes. Similar responses werefound in other dry tundra summer studies [Welker et al., 2000; Björkman et al., 2010] (Figure 7 and Table S1).However, our findings are in contrast to a previous study from our site [Rogers et al., 2011], where higher snowdepth enhanced Reco fluxes in the SFH, but not the SFM treatment. The difference between the two studiesmay be related to differences in weather conditions or possibly to changes in the substrate availability formicrobial degradation between 2007 and our measurement period. In 2007, the measurement period ofRogers et al. [2011], mean summer air temperature was 5.5 ± 3.8°C and thus similar to our study (5.3–6.1°C).However, water year precipitation was only 279±196mm and thus much less than in our study (410–557mm).Therefore, plants or microbes in the control and SFM plots might have been affected by drought stress that wasalleviated in the SFH treatment. In addition, since weather conditions are highly variable on a daily basis,difference between studies might derive from disparities in the number of measurements as Rogers et al. [2011]calculated Reco from about 8 to 10 points compared to our almost 50 measurements.

The dominant effect of an experimentally increased snowpack was a shorter growing season. As aconsequence, cumulative summer Reco was smaller in both experimental snowpack manipulations (SFM andSFH) compared to control levels (with the exception of SFH in 2011; Figure 3). This confirms earlier summerstudies in dry tundra systems [Björkman et al., 2010,Welker et al., 2000] where shorter plant growing seasonswere associated with deeper snowpack [Rieley et al., 1995; Cooper et al., 2011; Mallik et al., 2011; Semenchuket al., 2013].

Figure 7. Synthesis of mean daily summer ecosystem respiration (Reco) fluxes from wet (full symbols) and dry (open symbols)tundra in the (a) Low and (b) High Arctic under increased snowpack as a function of time since the experiment was established.Measurements from the same studies under ambient snowpack conditions are plotted at time zero. (Data from this studyand Björkman et al. [2010], Jones et al. [1999],Morgner et al. [2010], Natali et al. [2011], Olofsson et al. [2011], Rogers et al. [2011],andWelker et al. [2000]. See supporting information for details (Table S1).)



The timing and amount of future snow is uncertain and greatly varies between regions [Callaghan et al., 2011;Liston and Hiemstra, 2011]. Increases in erect shrub populations [Epstein et al., 2013; Pearson et al., 2013] willlocally create areas with shorter growing season, since shrubs act as natural snow barriers that cause snowaccumulation and delayedmelt [Sturm et al., 2005]. Whether changes in SCD and vegetation cover will have netpositive or negative climate feedbacks, effects remain uncertain today [Welker et al., 1997; Anderson-Smithet al., 2014].

Our study also demonstrates the effectiveness of long-term monitoring studies for understanding C cyclingin ecosystems with extreme interannual variability in weather such as polar semideserts. Compiling literaturedata from long-term snowpack manipulation studies in the High and Low Arctic showed supports ourhypothesis: Interannual variability of mean daily Reco within a given treatment was typically much larger thanthe treatment effect (Figure 7). Sustained, strong snow additions, however, significantly reduce the growingseason length and may result in water logging and disease and vegetation collapse [Hinkel and Hurd, 2006;Olofsson et al., 2011].

4.3. Soil CO2 Concentrations at Different Depths

Greater CO2 concentrations in the topsoil during the snowmelt period as opposed to later in the summersuggests that microbial activity was stimulated by the input of new water and/or C and nutrients frommelting snow leaching through the litter layer and surface soil [Hirano, 2005; Scott-Denton et al., 2006].In addition, some of this CO2 may originate from the release of previously produced CO2 that had beentrapped in the frozen active layer during the winter [Albert and Perron, 2000; Schimel et al., 2006].These short-term effects during the shoulder seasons may be especially important to quantify, whilemeasurements during midsummer only would fail to unravel the complexity of these High Arctic systems[Walker et al., 2008].

Long-term snowpack increases clearly affected belowground CO2 concentrations. Increased snowpackresulted in greater CO2 concentrations due to the additional melt water and greater SWC that persistedthroughout the growing seasons compared to ambient conditions in all years. These results are consistentwith a parallel study that found that weekly summer water additions increased belowground CO2concentrations [Lupascu et al., 2014a].

4.4. Radiocarbon Content of Ecosystem and Soil Respiration

Analyses of the 14C content of Reco and pore space CO2 provide a means to determine the sources ofecosystem C loss [e.g., Czimczik and Welker, 2010; Nowinski et al., 2010]. Ecosystem respiration is acombination of CO2 derived from three sources: (a) current plant metabolism (

14C ~30 per mil), (b)decomposition of recently fixed plant tissue (fixed since 1950: 14C >30 per mil), and (c) decompositionof old soil organic matter (14C

Soil CO2 was depleted (older) in14C under experimentally increased snowpack in both vegetated and bare

areas. Other studies from moist acidic tussock tundra in the Low Arctic found older C being emitted duringsummertime under increased snowpack manipulation in both Reco [Nowinski et al., 2010; Natali et al., 2011]and soil pore space [Natali et al., 2011]. Unfortunately, we were limited in our ability to collect samples tomeasure the 14C content of root respiration and microbial respiration in litter and bulk soil C data from theSFM and SFH treatments. More destructive sampling is required to predict the contributions of old Cdecomposition to Reco from soil CO2 measurements.

Together, older CO2 and higher CO2 concentrations suggest greater decomposition of soil C under deepersnow and contribute to increasing atmospheric CO2 concentrations and global climate change [Czimczik andWelker, 2010; Schaefer et al., 2011; Hicks-Pries et al., 2013; Lupascu et al., 2014a, 2014b]. In our monthlyobservations, however, we found no effect of snowpack on the mean age of Reco. This occurred in part,because Reco integrates C dynamics within the entire active layer as well as the aboveground plantcommunity. Emissions of old, deep Cwere likely masked by variable contributions from young, near-surface Cthat quickly respond to weather fluctuations. In addition, our inability to detect older C emissions from theexperimental treatments may be due in part to our discontinuous sampling scheme.

We observed episodic release of older CO2, irrespective of treatment. In June, some of the older values of Recoare likely due to the release of CO2 trapped over winter [Schimel et al., 2006]. In July and August, these

14Cdata further support episodic flushing of the active layer. The oldest values (most depleted in 14C) wemeasured all coincided with small intensity (8mm) coincided with observations of younger CO2 at depth (Figure 6d), suggesting that largerprecipitation events translocate modern C from the litter layer and/or rhizosphere to depth where it isdecomposed. The decomposition of this younger C is likely masking potential decomposition of older, in situC at these depths. These phenomena have been previously reported in an adjacent summertime climatemanipulation study [Lupascu et al., 2014a, 2014b].

5. Conclusions

Anticipated changes in the timing and amount of winter snowfall are uncertain, but our study revealed thecomplexity of how changes in snowfall regime can significantly affect C cycling in the High Arctic. Underexperimentally increased snowpack, we found a decrease in cumulative summertime Reco due to a reductionin growing season length. Interannually, the magnitude of summertime Reco showed a positive trend withwater year precipitation and thus SWC. In addition, our data suggested greater decomposition of older soilC under deeper snowpack. Additional, multiyear measurements of plant productivity and net ecosystemexchange along with more continuous 14C measurements of Reco and of soil C pools are needed toinvestigate the effects of snow depth on the total ecosystem C balance and land-atmosphere CO2 exchange.To quantify the annual C balance of High Arctic tundra, these studies should not cover only the growingseason but also the winter, as recent work demonstrated that microorganisms function well below formerlyassumed temperature and moisture limits.

ReferencesAdam, J. C., A. F. Hamlet, and D. P. Lettenmaier (2009), Implications of global climate change for snowmelt hydrology in the twenty-first

century, Hydrol. Process., 23, 962–972, doi:10.1002/hyp.Al-Kayssi, A. W., A. Al-Karaghouli, A. M. Hasson, and S. A. Beker (1990), Influence of soil moisture content on soil temperature and heat storage

under greenhouse conditions, J. Agr. Eng. Res., 45, 241–252.Albert, M. R., and F. E. Perron (2000), Ice layer and surface crust permeability in a seasonal snow pack, Hydrol. Process, 14, 3207–3214.Anderson-Smith, A., P. F. Sullivan, and J. M. Welker (2014), Increases in tundra shrub density is manifested in vegetation spectral properties

and corresponds to greater CO2 sequestration, Global Change Biol., in press.Bhatt, U. S., et al. (2010), Circumpolar arctic tundra vegetation change is linked to sea ice decline, Earth Interact., 14(8), 1–20, doi:10.1175/

2010EI315.1.Björkman, M. P., E. Morgner, R. G. Björk, E. J. Cooper, B. Elberling, and L. Klemedtsson (2010), A comparison of annual and seasonal carbon

dioxide effluxes between sub-Arctic Sweden and High-Arctic Svalbard, Polar Res., 29(1), 75–84, doi:10.1111/j.1751-8369.2010.00150.x.Borner, A. P., K. Kielland, and M. D. Walker (2008), Effects of simulated climate change on plant phenology and nitrogen mineralization in

Alaskan arctic tundra, Arct. Antarct. Alp. Res., 40(1), 27–38, doi:10.1657/1523-0430(06-099).Brooks, P. D., S. K. Schmidt, and M. W. Williams (1997), Winter production of CO2 and N2O from alpine tundra: Environmental controls and

relationship to inter-system C and N fluxes, Oecologia, 110, 403–413.Brooks, P. D., M. W. Williams, and S. K. Schmidt (1998), Inorganic nitrogen and microbial biomass dynamics before and during spring

snowmelt, Biogeochemistry, 43(1), 1–15.



AcknowledgmentsThis work was made possible by assis-tance from U.S. Air Force Base Thule,Greenland, CH2M Hill Polar Services, andthe KCCAMS laboratory. We thank A. Stills,J. Thomas, J. Yi, and M. Ruacho (UC Irvine),as well as C. Lett and K. Maseyk (UniversitéPierre et Marie Curie Paris 6, France) fortheir help in the field or laboratory.Furthermore, we thank J. Zautner(14

thWeather Squadron, Thule A.B.)

for sharing the climate data set, R.M.Nichols (UC Irvine) for assistance withthe statistics, and U. Seibt and W. Sun(UC, Los Angeles) for sharing theirexpertise on stable and radiocarbonisotopes modeling. This work wasfunded by the U.S. National ScienceFoundation (ARC-0909514 to C.I.C.and ARC-0909538 to J.M.W.).

http://dx.doi.org/10.1002/hyphttp://dx.doi.org/10.1175/2010EI315.1http://dx.doi.org/10.1175/2010EI315.1http://dx.doi.org/10.1111/j.1751&hyphen;8369.2010.00150.xhttp://dx.doi.org/10.1657/1523&hyphen;0430(06&hyphen;099)

Buckeridge, K. M., and P. Grogan (2010), Deepened snow increases late thaw biogeochemical pulses in mesic low arctic tundra,Biogeochemistry, 101, 105–121, doi:10.1007/s10533-010-9426-5.

Burnham, J. H., and R. S. Sletten (2010), Spatial distribution of soil organic carbon in northwest Greenland and underestimates of High Arcticcarbon stores, Global Biogeochem. Cycle, 24, GB3012, doi:10.1029/2009GB003660.

Callaghan, T. V., et al. (2011), Multiple effects of changes in arctic snow cover, Ambio, 40(S1), 32–45, doi:10.1007/s13280-011-0213-x.Chimner, R. A., and J. M. Welker (2005), Ecosystem respiration responses to experimental manipulations of winter and summer precipitation

in a Mixedgrass Prairie, WY, USA, Biogeochemistry, 73(1), 257–270, doi:10.1007/s10533-004-1989-6.Cooper, E. J., S. Dullinger, and P. Semenchuk (2011), Late snowmelt delays plant development and results in lower reproductive success in

the High Arctic, Plant Sci., 180(1), 157–167, doi:10.1016/j.plantsci.2010.09.005.Czimczik, C. I., and J. M. Welker (2010), Radiocarbon content of CO2 respired from High Arctic tundra in northwest Greenland, Arct. Antarct.

Alp. Res., 42(3), 342–350, doi:10.1657/1938-4246-42.3.342.Czimczik, C. I., S. E. Trumbore, M. S. Carbone, and G. C. Winston (2006), Changing sources of soil respiration with time since fire in a boreal

forest, Global Change Biol., 12(6), 957–971, doi:10.1111/j.1365-2486.2006.01107.x.De Michele, C., R. Vezzoli, H. Pavlopoulos, and R. J. Scholes (2008), A minimal model of soil water–vegetation interactions forced by stochastic

rainfall in water-limited ecosystems, Ecol. Model., 212(3-4), 397–407, doi:10.1016/j.ecolmodel.2007.10.035.Epstein, H. E., I. Myers-Smith, and D. A. Walker (2013), Recent dynamics of arctic and sub-arctic vegetation, Environ. Res. Lett., 8(1), 015040,

doi:10.1088/1748-9326/8/1/015040.Fahnestock, J. T., K. L. Povirk, and J. M. Welker (2000), Ecological significance of litter redistribution by wind and snow in arctic landscapes,

Ecography, 23, 623–631.Fahnestock, T., H. Jones, and M. Welker (1999), Wintertime CO2 efflux from arctic soils: Implications for annual carbon budgets, Global

Biogeochem. Cycle, 13(3), 775–779.Forbes, B. C., M. M. Fauria, and P. Zetterberg (2010), Russian Arctic warming and “greening” are closely tracked by tundra shrub willows,

Global Change Biol., 16(5), 1542–1554, doi:10.1111/j.1365-2486.2009.02047.x.Galen, C., and M. L. Stanton (1995), Responses of snowbed plant species to changes in growing-season length, Ecology, 76(5), 1546–1557.Gaudinski, J. B., S. E. Trumbore, A. Eric, and S. Zheng (2000), Soil carbon cycling in a temperate forest: Radiocarbon-based estimates of

residence times, sequestration rates and partitioning of fluxes, Biogeochemistry, 51, 33–69.Groendahl, L., T. Friborg, and H. Soegaard (2007), Temperature and snow-melt controls on interannual variability in carbon exchange in the

High Arctic, Theor. Appl. Climatol., 88(1-2), 111–125, doi:10.1007/s00704-005-0228-y.Hicks-Pries, C., E. A. G. Schuur, and K. G. Crummer (2013), Thawing permafrost increases old soil and autrotrophic respiration in tundra:

Partitioning ecosystem respiration using13C and Δ

14C, Global Change Biol., 19(2), 649–661, doi:10.1111/gcb.12058.

Hinkel, K. M., and J. K. Hurd (2006), Permafrost destabilization and thermokarst following snow fence installation, Barrow, Alaska, U.S.A., Arct.Antarct. Alp. Res., 38(4), 530–539.

Hinzman, L. D., et al. (2005), Evidence and implications of recent climate change in northern Alaska and other Arctic regions, Clim. Change,72(3), 251–298, doi:10.1007/s10584-005-5352-2.

Hirano, T. (2005), Seasonal and diurnal variations in topsoil and subsoil respiration under snowpack in a temperate deciduous forest, GlobalBiogeochem. Cycle, 19, GB2011, doi:10.1029/2004GB002259.

Horwath, J. L. (2007) Quantification and spatial assessment of High Arctic soil organic carbon storage in northwest Greenland, PhD thesis,Univ. Washington.

Horwath, J. L., R. S. Sletten, B. Hagedorn, and B. Hallet (2008), Spatial and temporal distribution of soil organic carbon in nonsorted stripedpatterned ground of the High Arctic, J. Geophys. Res., 113, G03S07, doi:10.1029/2007JG000511.

Jones, M. H., J. T. Fahnestock, D. A. Walker, M. D. Walker, and J. M. Welker (1998), Carbon dioxide fluxes in moist and dry arctic tundra during thesnow-free season: Responses to increases in summer temperature and winter snow accumuation, Arct. Antarct. Alp. Res., 30(4), 373–380.

Jones, M. H., J. T. Fahnestock, and J. M. Welker (1999), Early and late winter CO2 flux from arctic tundra in the Kuparuk River watershed, Alaska,U.S.A., Arct. Antarct. Alp. Res., 31(2), 187–190.

Jones, M. H., J. T. Fahnestock, P. D. Stahl, and J. M. Welker (2000), A note on summer CO2 flux, soil organic and microbial biomass from dif-ferent High Arctic ecosystems types in northwestern Greenland, Arct. Antarct. Alp. Res., 32(1), 104–106.

Kessler, M. A., and B. T. Werner (2003), Self-organization of sorted patterned ground, Science, 299, 380–383, doi:10.1126/science.1077309.Lafrenière, M. J., E. Laurin, and S. F. Lamoureux (2013), Impact of snow augmentation on solute fluxes from High Arctic headwater catch-

ments, Vadose Zone J. 12, vzj2012.0058, doi:10.5194/bg-8-1193-2011.Lawrence, D. M., A. G. Slater, R. A. Tomas, M. M. Holland, and C. Deser (2008), Accelerated Arctic land warming and permafrost degradation

during rapid sea ice loss, Geophys. Res. Lett., 35, L11506, doi:10.1029/2008GL033985.Leffler, A. J., and J. M. Welker (2013), Long-term increases in snow pack elevate leaf N and photosynthesis in Salix arctica: Responses to a

snow fence experiment in the High Arctic of NW Greenland, Environ. Res. Lett., 8(2), 025023, doi:10.1088/1748-9326/8/2/025023.Levin, I., K. O. Munnich, and W. Weiss (1980), The effect of anthropogenic CO2 and 14C sources on the distribution of 14C in the atmosphere,

Radiocarbon, 22(2), 379–391.Liston, G. E., and C. A. Hiemstra (2011), The changing cryosphere: Pan-Arctic snow treds (1979–2009), J. Clim. Res., 24(21), 5691–5712,

doi:10.1175/JCLI-D-11-00081.1.Lloyd, C. R. (2001), On the physical controls of the carbon dioxide balance at a High Arctic site in Svalbard, Theor. Appl. Climatol., 70(1-4),

167–182, doi:10.1007/s007040170013.Lloyd, J., and J. A. Taylor (1994), On the temperature dependence of soil respiration, Funct. Ecol., 8(3), 315–323, doi:10.2307/2389824.Lupascu, M., J. M. Welker, U. Seibt, K. Maseyk, X. Xu, and C. I. Czimczik (2014a), Arctic wetting dampens permafrost carbon feedbacks to

climate warming, Nat. Clim. Change, 4, 51–55, doi:10.1038/nclimate2058.Lupascu, M., J. M. Welker, U. Seibt, X. Xu, I. Velicogna, D. S. Lindsey, and C. I. Czimczik (2014b), Water availability regulates magnitude,

seasonality and sources of ecosystem respiration from a polar semi-desert in NW Greenland, Biogeosciences Discuss., 11, 1–40,doi:10.5194/bgd-11-1-2014.

Mallik, A. U., J. V. Wdowiak, and E. J. Cooper (2011), Growth and reproductive responses of Cassiope tetragona, a circumpolar evergreenshrub, to experimentally delayed snowmelt, Arct. Antarct. Alp. Res., 43(3), 404–409, doi:10.1657/1938-4246-43.3.404.

Molau, U., and S. Edlund (1996), Plant response variables, ITEX, 23–34.Morgner, E., B. Elberling, D. Strebel, and E. J. Cooper (2010), The importance of winter in annual ecosystem respiration in the High Arctic:

Effects of snow depth in two vegetation types, Polar Res., 29(1), 58–74, doi:10.1111/j.1751-8369.2010.00151.x.Natali, S. M., E. A. G. Schuur, C. Trucco, C. E. Hicks Pries, K. G. Crummer, and A. F. Baron Lopez (2011), Effects of experimental warming of air, soil

and permafrost on carbon balance in Alaskan tundra, Global Change Biol., 17(3), 1394–1407, doi:10.1111/j.1365-2486.2010.02303.x.



http://dx.doi.org/10.1007/s10533&hyphen;010&hyphen;9426&hyphen;5http://dx.doi.org/10.1029/2009GB003660http://dx.doi.org/10.1007/s13280&hyphen;011&hyphen;0213&hyphen;xhttp://dx.doi.org/10.1007/s10533&hyphen;004&hyphen;1989&hyphen;6http://dx.doi.org/10.1016/j.plantsci.2010.09.005http://dx.doi.org/10.1657/1938&hyphen;4246&hyphen;42.3.342http://dx.doi.org/10.1111/j.1365&hyphen;2486.2006.01107.xhttp://dx.doi.org/10.1016/j.ecolmodel.2007.10.035http://dx.doi.org/10.1088/1748&hyphen;9326/8/1/015040http://dx.doi.org/10.1111/j.1365&hyphen;2486.2009.02047.xhttp://dx.doi.org/10.1007/s00704&hyphen;005&hyphen;0228&hyphen;yhttp://dx.doi.org/10.1111/gcb.12058http://dx.doi.org/10.1007/s10584&hyphen;005&hyphen;5352&hyphen;2http://dx.doi.org/10.1029/2004GB002259http://dx.doi.org/10.1029/2007JG000511http://dx.doi.org/10.1126/science.1077309http://dx.doi.org/10.5194/bg&hyphen;8&hyphen;1193&hyphen;2011http://dx.doi.org/10.1029/2008GL033985http://dx.doi.org/10.1088/1748&hyphen;9326/8/2/025023http://dx.doi.org/10.1175/JCLI&hyphen;D&hyphen;11&hyphen;00081.1http://dx.doi.org/10.1007/s007040170013http://dx.doi.org/10.2307/2389824http://dx.doi.org/10.1038/nclimate2058http://dx.doi.org/10.5194/bgd&hyphen;11&hyphen;1&hyphen;2014http://dx.doi.org/10.1657/1938&hyphen;4246&hyphen;43.3.404http://dx.doi.org/10.1111/j.1751&hyphen;8369.2010.00151.xhttp://dx.doi.org/10.1111/j.1365&hyphen;2486.2010.02303.x

Nobrega, S., and P. Grogan (2013), Deeper snow enhances winter respiration from both plant-associated and bulk soil carbon pools in birchhummock tundra, Ecosystems, 10(3), 419–431, doi:10.1007/sl0021-007-9033-z.

Nowinski, N. S., L. Taneva, S. E. Trumbore, and J. M. Welker (2010), Decomposition of old organic matter as a result of deeper active layers in asnow depth manipulation experiment, Oecologia, 163(3), 785–92, doi:10.1007/s00442-009-1556-x.

Olofsson, J., L. Ericson, M. Torp, S. Stark, and R. Baxter (2011), Carbon balance of Arctic tundra under increased snow cover mediated by aplant pathogen, Nat. Clim. Change, 1(4), 220–223, doi:10.1038/nclimate1142.

Overpeck, J., et al. (1997), Arctic environmental change of the last four centuries, Science, 278(5341), 1251–1256.Pearson, R. G., S. J. Phillips, M. M. Loranthy, P. S. A. Beck, T. D. Damoulas, S. J. Knight, and S. J. Goetz (2013), Shifts in Arctic vegetation and

associated feedbacks under climate change, Nat. Clim. Change, 3, 673–677, doi:10.1038/nclimate1858.Phillips, C. L., K. J. McFarlane, D. Risk, and A. R. Desai (2013), Biological and physical influences on soil

14CO2 seasonal dynamics in a temperate

hardwood forest, Biogeosciences, 10, 7999–8012.Pomeroy, J. W., D. S. Bewley, R. L. H. Essery, N. R. Hedstrom, T. Link, R. J. Granger, J. E. Sicart, C. R. Ellis, and J. R. Janowicz (2006), Shrub tundra

snowmelt, Hydrol. Process., 20(4), 923–941, doi:10.1002/hyp.6124.Rieley, G., J. M. Welker, G. Eglinton, and T. V. Callaghan (1995), Epicuticular waxes of arctic plants: Compositional differences in relation to

winter snow cover, Phytochem, 38, 45–52.Rogers, M. C., P. F. Sullivan, and J. M. Welker (2011), Evidence of nonlinearity in the response of net ecosystem CO2 exchange to increasing levels

of winter snow depth in the High Arctic of northwest Greenland, Arct. Antarct. Alp. Res., 43(1), 95–106, doi:10.1657/1938-4246-43.1.95.Schaefer, K., T. Zhang, L. Bruhwiler, and A. P. Barrett (2011), Amount and timing of permafrost carbon release in response to climate warming,

Tellus B, 63(2), 165–180, doi:10.1111/j.1600-0889.2011.00527.x.Schimel, J. P., C. Bilbrough, and J. M. Welker (2004), Increased snow depth affects microbial activity and nitrogen mineralization in two Arctic

tundra communities, Soil Biol. Biochem., 36(2), 217–227, doi:10.1016/j.soilbio.2003.09.008.Schimel, J. P., J. Fahnestock, G. Michaelson, C. Mikan, C. Ping, E. Romanovsky, and J. Welker (2006), Cold-season production of CO2 in Arctic soils:

Can laboratory and field estimates be reconciled through a simple modeling approach?, Arct. Antarct. Alp. Res., 38(2), 249–256.Schuur, E. A. G., and S. E. Trumbore (2006), Partitioning sources of soil respiration in boreal black spruce forest using radiocarbon, Global

Change Biol., 12(2), 165–176, doi:10.1111/j.1365-2486.2005.01066.x.Scott-Denton, L. E., T. N. Rosenstiel, and R. K. Monson (2006), Differential controls by climate and substrate over the heterotrophic and rhi-

zospheric components of soil respiration, Global Change Biol., 12(2), 205–216, doi:10.1111/j.1365-2486.2005.01064.x.Semenchuk, P. R., B. Elberling, and E. J. Cooper (2013), Snow cover and extreme winter warming events control flower abundance of some,

but not all species in High Arctic Svalbard, Ecol. Evol., 3(8), 2586–2599, doi:10.1002/ece3.648.Sharp, E. D., P. F. Sullivan, H. Steltzer, and A. Z. Csank (2013), Complex carbon cycle responses to multi-level warming and supplemental

summer rain in the High Arctic, Global Change Biol., 1–13, doi:10.1111/gcb.12149.Southon, J., and G. M. Santos (2007), Life with MC-SNICS. Part II: Further ion source development at the Keck carbon cycle AMS facility, Nucl.

Instrum. Methods Phys. Res., Sect. B., 259(1), 88–93, doi:10.1016/j.nimb.2007.01.147.Steffen, W. (2006), The Arctic in an earth system context: From brake to accelerator of change, Ambio, 35(4), 153–9.Stuiver, M., and H. A. Polach (1977), Radiocarbon, Radiocarbon, 19(3), 355–363.Sturm, M., J. Schimel, G. Michaelson, J. M. Welker, S. F. Oberbauer, G. E. Liston, J. Fahnestock, and V. E. Romanovsky (2005), Winter biological

processes could help convert arctic tundra to shrubland, BioScience, 55(1), 17–26, doi:10.1641/0006-3568(2005)055[0017:WBPCHC]2.0.CO;2.Sullivan, P. F., and J. M. Welker (2007), Variation in leaf physiology of Salix arctica within and across ecosystems in the High Arctic: Test of a

dual isotope conceptual model, Oecologia, 151, 372–386.Sullivan, P. F., J. M. Welker, H. Steltzer, R. S. Sletten, B. Hagedorn, S. J. T. Arens, and J. L. Horwath (2008), Energy and water additions give rise to

simple responses in plant canopy and soil microclimates of a High Arctic ecosystem, J. Geophys. Res., 113, G03S08, doi:10.1029/2007JG000477.Tape, K., M. Sturm, and C. Racine (2006), The evidence for shrub expansion in northern Alaska and the Pan-Arctic, Global Change Biol., 12(4),

686–702, doi:10.1111/j.1365-2486.2006.01128.x.Torn, M. S., M. Kleber, E. S. Zavaleta, B. Zhu, C. B. Field, and S. E. Trumbore (2013), A dual isotope approach to isolate soil carbon pools of

different turnover times, Biogeosciences, 10, 8067–8081, doi:10.5194/bg-10-8067-2013.Trumbore, S. (2006), Carbon respired by terrestrial ecosystems—Recent progress and challenges, Global Change Biol., 12, 141–153,

doi:10.1111/j.1365-2486.2005.01067.x.Trumbore, S. (2009), Radiocarbon and soil carbon dynamics, Ann. Rev. Earth Pl. Sc., 37(1), 47–66, doi:10.1146/annurev.

earth.36.031207.124300.U.S. Department of Agriculture (1999), Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys, U.S.

Government Printing Office, Washington, D. C.Wahren, C.-H. A., M. D. Walker, and M. S. Bret-Harte (2005), Vegetation responses in Alaskan arctic tundra after 8 years of a summer warming

and winter snow manipulation experiment, Global Change Biol., 11(4), 537–552, doi:10.1111/j.1365-2486.2005.00927.x.Walker, D. A., H. E. Epstein, and J. M. Welker (2008), Introduction to the special section: Biocomplexity in Arctic terrestrial environments,

J. Geophys. Res., 113, G03S00, doi:10.1029/2008JG000740.Walker, D. A., et al. (2004), Frost-boil ecosystems: Complex interactions between landforms, soils, vegetation and climate, Permafrost Periglac.

Process., 15(2), 171–188, doi:10.1002/ppp.487.Walker, M. D., et al. (1999), Long-term experimental manipulation of winter snow regime and summer temperature in arctic and alpine

tundra, Hydrol. Process., 13, 2315–2330.Welker, J. M., J. T. Fahnestock, and M. H. Jones (2000), Annual CO2 flux in dry and moist arctic tundra: Field responses to increases in summer

temperatures and winter snow depth, Clim. Change, 44, 139–150.Welker, J. M., J. T. Fahnestock, G. H. R. Henry, K. W. O’Dea, and R. A. Chimner (2004), CO2 exchange in three Canadian High Arctic ecosystems:

Response to long-term experimental warming, Global Change Biol., 10(12), 1981–1995, doi:10.1111/j.1365-2486.2004.00857.x.Welker, J. M., J. T. Fahnestock, P. F. Sullivan, and R. A. Chimner (2005), Leaf mineral nutrition of Arctic plants in response to warming and

deeper snow in northern Alaska, Oikos, 109, 167–177.Welker, J. M., U. Molau, A. N. Parsons, C. Robinson, and P. A. Wookey (1997), Response of Dryas octopetala to ITEX manipulations: A synthesis

with circumpolar comparisons, Global Change Biol., 3(Suppl. 1), 61–73.Woo, M., R. Heron, P. Marsh, and P. Steer (1983), Comparison of weather station snowfall with winter snow accumulation in High Arctic

basins, Atmos. Ocean, 21(3), 312–325, doi:10.1080/07055900.1983.9649171.Xu, X., S. E. Trumbore, S. Zheng, J. R. Southon, K. E. McDuffee, M. Luttgen, and J. C. Liu (2007), Modifying a sealed tube zinc reduction method

for preparation of AMS graphite targets: Reducing background and attaining high precision, Nucl. Instrum. Methods Phys. Res., Sect. B,259(1), 320–329, doi:10.1016/j.nimb.2007.01.175.



http://dx.doi.org/10.1007/sl0021&hyphen;007&hyphen;9033&hyphen;zhttp://dx.doi.org/10.1007/s00442&hyphen;009&hyphen;1556&hyphen;xhttp://dx.doi.org/10.1038/nclimate1142http://dx.doi.org/10.1038/nclimate1858http://dx.doi.org/10.1002/hyp.6124http://dx.doi.org/10.1657/1938&hyphen;4246&hyphen;43.1.95http://dx.doi.org/10.1111/j.1600&hyphen;0889.2011.00527.xhttp://dx.doi.org/10.1016/j.soilbio.2003.09.008http://dx.doi.org/10.1111/j.1365&hyphen;2486.2005.01066.xhttp://dx.doi.org/10.1111/j.1365&hyphen;2486.2005.01064.xhttp://dx.doi.org/10.1002/ece3.648http://dx.doi.org/10.1111/gcb.12149http://dx.doi.org/10.1016/j.nimb.2007.01.147http://dx.doi.org/10.1641/0006&hyphen;3568(2005)055[0017:WBPCHC]2.0.CO;2http://dx.doi.org/10.1029/2007JG000477http://dx.doi.org/10.1111/j.1365&hyphen;2486.2006.01128.xhttp://dx.doi.org/10.5194/bg&hyphen;10&hyphen;8067&hyphen;2013http://dx.doi.org/10.1111/j.1365&hyphen;2486.2005.01067.xhttp://dx.doi.org/10.1146/annurev.earth.36.031207.124300http://dx.doi.org/10.1146/annurev.earth.36.031207.124300http://dx.doi.org/10.1111/j.1365&hyphen;2486.2005.00927.xhttp://dx.doi.org/10.1029/2008JG000740http://dx.doi.org/10.1002/ppp.487http://dx.doi.org/10.1111/j.1365&hyphen;2486.2004.00857.xhttp://dx.doi.org/10.1080/07055900.1983.9649171http://dx.doi.org/10.1016/j.nimb.2007.01.175

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.00000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 400 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.00000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects true /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > > /FormElements true /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles true /MarksOffset 6 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /UseDocumentProfile /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice

rates and radiocarbon content of summer ecosystem ... · the study period (2010, 2011, and 2012)...

Documents